G-212: Find Incorrect Arrow Direction Data Generator

Generates synthetic tasks where multiple arrows are arranged in a circle. The goal is to identify the arrow pointing in a different direction from all others and circle it.

Each sample pairs a task (first frame + prompt describing what needs to happen) with its ground truth solution (final frame showing the result + video demonstrating how to achieve it). This structure enables both model evaluation and training.

---

📌 Basic Information

Property	Value
Task ID	G-212
Task	Find Incorrect Arrow Direction
Category	Perception
Resolution	1024×1024 px
FPS	16 fps
Duration	~3 seconds
Output	PNG images + MP4 video

---

🚀 Usage

Installation

# Clone the repository
git clone https://github.com/VBVR-DataFactory/G-212_find_incorrect_arrow_direction_data-generator.git
cd G-212_find_incorrect_arrow_direction_data-generator

# Install dependencies
pip install -r requirements.txt

Generate Data

# Generate 100 samples
python examples/generate.py --num-samples 100

# Generate with specific seed
python examples/generate.py --num-samples 100 --seed 42

# Generate without videos
python examples/generate.py --num-samples 100 --no-videos

# Custom output directory
python examples/generate.py --num-samples 100 --output data/my_output

Command-Line Options

Argument	Type	Description	Default
--num-samples	int	Number of samples to generate	100
--seed	int	Random seed for reproducibility	Random
--output	str	Output directory	data/questions
--no-videos	flag	Skip video generation	False

---

📖 Task Example

Prompt

Look at the arrows arranged in a circle on the screen. Please circle the arrow pointing in a different direction from the others.

Visual

Initial Frame Arrows arranged in circle	Animation Red circle appears around incorrect arrow	Final Frame Incorrect arrow circled

---

📖 Task Description

Objective

Identify and circle the single arrow that points in a different direction from all other arrows in a circular arrangement.

Task Setup

• Arrow Count: 4-14 arrows per image
• Arrangement: Arrows evenly spaced in a circular pattern (no visible reference circle)
• Arrow Types: All arrows identical in size and style
• Target: Exactly one arrow pointing in different direction (the outlier)
• Marking Method: Red circle appears around the incorrect arrow
• Radius Variation: Circle radius varies (0.28-0.36 relative to image size) for sample diversity
• Visual Style: Clean white background with arrows only (no reference circle for cleaner appearance)

Key Features

• Pattern recognition: Tests ability to identify anomalies in uniform arrangements
• Direction comparison: Requires comparing arrow orientations across the display
• Outlier detection: Finding the single different element among uniform ones
• Circular layout: Spatial organization in circular pattern
• Clear animation: Red circle smoothly appears around target
• Unambiguous answer: Only one arrow differs per task instance

---

📦 Data Format

data/questions/find_incorrect_arrow_direction_task/find_incorrect_arrow_direction_00000000/
├── first_frame.png      # Initial state (arrows without marking)
├── final_frame.png      # Goal state (incorrect arrow circled)
├── prompt.txt           # Task instructions
├── ground_truth.mp4     # Solution video (16 fps)
└── question_metadata.json # Task metadata

File specifications: Images are 1024×1024 PNG. Videos are MP4 at 16 fps, approximately 3 seconds long.

---

🏷️ Tags

pattern-recognition visual-search anomaly-detection direction-comparison circular-arrangement outlier-detection

---